1) Field of the Invention
The present invention relates to a technology for recognizing dynamically a change in route conditions in a network and selecting an optimum route for transmission of data.
2) Description of the Related Art
In a network, when a device transmits data to another device, the data passes through a plurality of nodes. The nodes may be transmission apparatuses, switchboards, switches, routers, wavelength division multiplexing (WDM) apparatuses, asynchronous transfer mode (ATM) apparatuses, optical cross-connectors, and fiber channels.
Depending on the topology of the network, there may be numerous possible routes for the data. Along the path to the destination, each node needs to select a route for transmitting the data further. Generally, a communication cost between nodes is set in advance, and the nodes select the route based on the communication cost set.
The communication cost is a value, which is an index of the condition of the communication line connecting the nodes. Conventionally, the communication cost in each communication line is set empirically based on factors such as the line speed of the communication line. For example, the greater the priority level of a communication line, the lower is its communication cost. Moreover, the higher is the line speed, the lower is its communication cost.
Each route to the destination includes a plurality of communication lines. In the conventional communication apparatus, a sum total of the communication costs of the communication lines included in a route is calculated. The communication apparatus selects the route having the lowest value of the communication cost, so that the data can be transmitted to the destination efficiently and quickly. The communication apparatus serves as a node.
There are various methods of selecting the route of transmission using the communication cost. A network design method in Japanese Patent Application Laid-open No. H10-207934 discloses a method for solving a mathematical plan problem. This network design method assumes a plurality of routes to the destination of data, sets a restrictive condition such that the communication cost in each route does not exceed a predetermined value, and specifies an optimum route by finding out an objective function with the lowest communication cost of the route.
In a communication network design apparatus disclosed in Japanese Patent Application Laid-open No. 2000-165451, a plurality of nodes are arranged to optimize the topology of the network, taking into consideration the maximum line speed and the minimum line speed between nodes, and a preset communication cost is used for determining whether the arrangement of the respective nodes is optimum.
However, there are some problems in applying methods involving communication cost, to the conventional apparatus. Once a communication cost between respective nodes is set, the same communication cost is used, unless a network administrator changes the value. Moreover, because the quantity of data that passes through the communication line is not always constant, the line speed of the communication line may be used to its maximum, or may not be used at all. Furthermore, the service condition of a communication line changes with the lapse of time. In such a case, the communication cost of that communication line must change accordingly. However, in the conventional apparatus, the communication cost can be corrected only by the network administrator on recognizing the failure. Therefore, the nodes may try to use the faulty communication line, and hence smooth data transmission is not possible.
Thus, in the conventional communication apparatus, there is a problem in that the preset communication cost of the communication line may not always correspond to the actual condition of the communication line. In fact, it is very important to allow the communication cost to dynamically correspond to a change in the condition of the communication line to ensure that data reaches the destination quickly and without error.